Leverage Causes Fat Tails and Clustered Volatility...

A short, excellent paper by Stefan Thurner, J. Doyne Farmer and John Geanakoplos entiteled: "Leverage Causes Fat Tails and Clustered Volatility."

As noted by the authors, previous literature on endogenous processes that lead to fat tails and clustered volatility in distributions of asset price fluctuations have focused on various forms of "irrational trading behavior" as the root cause.  This paper demonstrates how the ability of traders to buy more assets than there weath would allow by borrowing money from banks (i.e., leverage) causes asset price flucatuations to exhibit fat tails and clustered volatility. 

In their model one of the causes of systemic risk is perversely the risk control policies of the banks themselves.  A prudent bank will seek to insure itself against by placing limits on the amount of leverage that a trader can maintain.  When a trader exceeds his leverage limit he is forced to repay some of his loan by selling the underlying asset.  Crashes can arise when a large number of traders find themselves in excess of their leverage limits (and thus having to repay) when the price of the underlying asset is already falling.  The resulting non-linear positive feedback reinforces the negative price movement causing it to fall even further, etc.  Now in the real world banks often seek to adjust leverage limits based on recent market conditions (i.e., banks will raise leverage limits when the market has been relatively stable, and will lower leverage limits when markets have been highly volatile).  Based on the linkages shown in the model, this would only make matters worse as traders would be deprived of funds at the moment they need them the most.

Major takeaways for me:

1. The linkage between leverage, fat-tails, and clustered volatility in the distribution of asset price fluctuations.
2. Sensible risk management policy at the individual bank level, can easily lead to increasing levels of risk exposure for the banking system as a whole. Sensible regulatory policy aimed at mitigating systemic risk should not rely soley on mitigating risk at the level of the individual bank.
3. This does not imply that leverage limits are a bad thing!

Perhaps what is needed is substantial diversity of leverage limits in the banking system, as opposed to a set of rules that apply to all banks?  I am reading "The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies" by Scott Page at the moment, and thus the potential benefits of diversity are fresh in my mind...